Vonage fraud protection with network API and generative AI | Amazon Web Services

Summary

The video script introduces Fraud Defender, a solution by Vonage, which leverages AWS AI and generative AI to combat identity theft and telecommunications fraud. It discusses challenges faced by organizations and details a cloud-native, serverless architecture that enhances security without adding user friction. The system uses Amazon Bedrock for feedback on transactions, silent authentication, and Amazon Recognition for facial verification, integrating network APIs for real-time insights to protect against fraud while making the authentication process smarter and more secure.

Takeaways

• 🌟 Vonage is partnering with AWS to introduce Fraud Defender, an innovative solution leveraging AWS AI/ML and generative AI technologies.
• 🛡️ The primary goal of Fraud Defender is to tackle challenges in identity theft, account takeover, and telecommunications fraud within the financial sector.
• 💡 Vonage's solution uses a cloud-native, serverless architecture deployed on AWS, offering customization and ease of purchase through the AWS Marketplace.
• 🔍 Fraud protection is enhanced by real-time network insights provided by Vonage's APIs, which are based on the Camara Network APIs.
• 🤖 Amazon Bedrock, a generative AI and large language model (LLM), is utilized to analyze transactions and provide feedback on their legitimacy.
• 🔄 The system creates a feedback loop, using insights from Amazon Bedrock to refine Fraud Defender's responses to fraudulent activities.
• 📞 Phone number verification is a key component, checking against Fraud Defender's knowledge base to identify potentially fraudulent numbers.
• 📍 Device location verification ensures that the user's claimed location matches the actual location data from telecom networks, adding a layer of security.
• 🔑 Silent authentication offers a secure method to replace one-time passwords, reducing the risk of social engineering attacks.
• 👤 Amazon Recognition uses facial recognition for user authentication, including a liveness check to prevent the use of stolen photos or videos.
• 🔄 The integration of generative AI allows for continuous improvement of security measures, adapting to new threats and maintaining an optimal balance between user convenience and fraud prevention.

Q & A

What is the role of Chad Hrin in the context of the script?

Chad Hrin is the Principal Solutions Architect with AWS, focusing on Communications as a Service (CaaS) and generative AI. He introduces the partnership with Vonage and their solution, Fraud Defender.

What is Vonage known for in the context of the script?

Vonage is known for being an innovative and progressive partner that offers Communication Platform as a Service (CPaaS) solutions and has developed Fraud Defender, a solution that leverages AWS AI and ML technologies.

What challenges does Vonage's Fraud Protection Solution address?

The solution addresses challenges related to identity theft, account takeover, and telecommunications fraud, which can result in financial losses and brand damage for organizations.

What is the significance of generative AI in the Fraud Defender solution?

Generative AI plays a crucial role in Fraud Defender by providing real-time insights and feedback on transactions, helping to identify fraudulent activities and improve the system's response to such threats over time.

How does the architecture of the Fraud Defender solution benefit customers?

The architecture is cloud-native and serverless, deployed on AWS, allowing customers to purchase through the AWS Marketplace and benefit from its customization and advantages.

What is Amazon Bedrock, and how does it relate to Fraud Defender?

Amazon Bedrock is a generative AI and large language model (LLM) that provides feedback on transactions to determine if they were fraudulent. This feedback is used to enhance Fraud Defender's capabilities.

What is the purpose of the silent authentication process in Fraud Defender?

Silent authentication is a secure process that can replace one-time passwords, preventing social engineering and providing a better guarantee of the phone line used for user authentication.

How does Amazon Recognition contribute to the user authentication process?

Amazon Recognition contributes by performing user authentication based on facial recognition, including a liveness check to prevent fraudsters from using stolen photos or videos for authentication.

What are the three Network APIs mentioned in the script, and what do they provide?

The three Network APIs are number verification, SIM swap, and verify location. They provide real-time insights into the telecommunications network, helping to identify fraudulent activities.

How does the Fraud Defender solution reduce user friction while increasing security?

By using mechanisms like silent authentication and facial recognition, Fraud Defender reduces user friction while increasing security only when necessary, based on insights indicating a potential fraudulent attempt.

What is the long-term impact of using generative AI in Fraud Defender?

The use of generative AI in Fraud Defender allows the system to become smarter, more intelligent, and more predictive over time, leading to increased security and fine-tuned parameters that match the perceived risk on the network.

Outlines

00:00

🤖 Introduction to Vonage's Fraud Defender

The first paragraph introduces Chad Hrin, an AWS principal solutions architect, who focuses on Communications as a Service (CaaS) and generative AI. He presents Vonage, a partner of AWS, and their innovative solution called Fraud Defender. Vonage offers a Communication Platform as a Service (CPaaS) and uses AWS AI and ML to combat fraud. The paragraph also introduces Bort from Vonage, who will walk through the value of this industry-first solution, addressing challenges in identity theft, account takeover, and telecommunications fraud. The solution aims to protect organizations from financial losses and brand damage caused by these fraudulent activities.

05:00

🛡️ Architecture and Capabilities of Fraud Defender

The second paragraph delves into the architecture of Vonage's Fraud Defender, which is a cloud-native, serverless solution deployed on AWS. Customers can purchase it through the AWS Marketplace and benefit from its customizable nature. The architecture includes an application for user interaction with financial services, an application backend, and a suite of APIs based on the Camara Network. These APIs provide real-time insights to identify fraudulent transactions or authentication processes. The collected data feeds into Amazon Bedrock, a generative AI and ML model, which provides feedback on transactions to enhance Fraud Defender's capabilities. The system closes the loop between real-time events and the system's response to fraudulent attacks, ensuring both security and user accessibility.

10:02

🚀 Enhancing Security with Network APIs and AI

The final paragraph discusses the advanced features of Fraud Defender, including network APIs for number verification, SIM swap checks, and location verification. These APIs provide real-time insights that enhance the system's ability to detect fraud. The paragraph also highlights silent authentication and facial recognition via Amazon Recognition, which improve security while reducing user friction. The generative AI component of Fraud Defender learns from insights to increase security over time, fine-tuning parameters to match the perceived risk on the network. The partnership with AWS and the integration of network APIs and AI make Fraud Defender a cutting-edge solution in the marketplace, setting the stage for future innovations.

Keywords

💡AWS

AWS stands for Amazon Web Services, which is a subsidiary of Amazon that provides on-demand cloud computing platforms and APIs to individuals, companies, and governments. In the video's context, AWS is highlighted as the platform where Vonage's Fraud Defender solution is deployed, showcasing the integration of AWS's AI and ML capabilities within the solution.

💡Communications as a Service (CaaS)

Communications as a Service refers to the delivery of communication services over the internet, allowing businesses to use these services without the need for on-premises hardware. In the script, Chad Hrin mentions his focus on CaaS, indicating the importance of cloud-based communication services in the development of innovative solutions like Fraud Defender.

💡Generative AI

Generative AI refers to artificial intelligence systems that can generate new content, such as text, images, or audio, based on existing data. The video discusses how Vonage's Fraud Defender leverages generative AI, particularly through Amazon Bedrock, to analyze transactions and improve fraud detection capabilities.

💡Fraud Defender

Fraud Defender is a solution offered by Vonage that focuses on combating fraud and identity theft within communication platforms. The script emphasizes how Fraud Defender integrates with AWS services and uses AI to protect against fraudulent activities, such as account takeover and telecommunications fraud.

💡Identity Theft

Identity theft is the unauthorized use of someone's personal identifying information, often to commit fraud. The video script discusses how organizations face challenges with identity theft, and how Fraud Defender helps in mitigating such risks by combining telecommunications capabilities with AI to detect and prevent unauthorized access to customer accounts.

💡Account Takeover

Account takeover refers to a type of cyber attack where a fraudster gains control over a user's account, often leading to financial losses and brand damage. The script mentions account takeover as a significant challenge that Fraud Defender addresses by using real-time network insights and AI to identify and prevent such attacks.

💡Telecommunications Fraud

Telecommunications fraud involves the misuse of telecommunication services, such as phone lines or text messages, to commit fraud. The script explains how Fraud Defender helps protect against this by detecting fraudulent use of verification processes and preventing financial exploitation.

💡Cloud Native and Serverless Architecture

A cloud-native and serverless architecture refers to a system designed to run on cloud services, without the need for managing servers. The script describes the architecture of Fraud Defender as being cloud-native and serverless, deployed on AWS, which allows for scalability, flexibility, and ease of customization.

💡Amazon Bedrock

Amazon Bedrock, as mentioned in the script, is a generative AI and large language model (LLM) that provides feedback on transactions to determine if they are fraudulent. It is part of the Fraud Defender solution's architecture, enhancing the system's ability to learn and adapt to new fraud patterns.

💡Silent Authentication

Silent authentication is a security process that verifies the identity of a user without their active participation, often using behavioral biometrics or other passive methods. The script highlights silent authentication as a feature of Fraud Defender that increases security by replacing one-time passwords and preventing social engineering attacks.

💡Amazon Recognition

Amazon Recognition is a service that provides user authentication through facial recognition, including a liveness check to ensure the authenticity of the user. The script discusses how Amazon Recognition is integrated into Fraud Defender to add an extra layer of security, making it harder for fraudsters to use stolen photos or videos for authentication.

Highlights

Introduction of Chad Hrin, Principal Solutions Architect with AWS, focusing on Communications as a Service (CX) and generative AI.

Vonage, an innovative partner, introduces Fraud Defender, a solution leveraging AWS AI/ML and generative AI.

Fraud Defender addresses challenges in identity theft and account takeover in the financial sector.

Telecommunications fraud prevention, protecting against fraudulent verification processes.

The architecture of Fraud Defender is cloud-native and serverless, deployed on AWS.

Customers can purchase Fraud Defender through AWS Marketplace for easy customization.

Real-time network insights are provided by Vonage APIs based on the Camara Network APIs.

Amazon Bedrock, a generative AI and LLM model, is used for transaction feedback to identify fraud.

Fraud Defender uses feedback to close the loop on real-time fraudulent attack responses.

Multi-layer fraud detection process including phone number verification and telecom network checks.

Sim swap checks to prevent telecom fraud and account takeover attempts.

Silent authentication as a secure process replacing one-time passwords and preventing social engineering.

Device location verification using real-time network information to ensure user location authenticity.

Amazon Recognition for user authentication with facial recognition and liveliness checks.

Network APIs discussed include number verification, Sim Swap, and verify location.

Fraud protection increases without adding user friction through mechanisms like silent authentication.

Amazon Recognition and generative AI enhance product intelligence, security, and predictive capabilities.

Fraud Defender's generative AI closes the loop between authentication and fraudulent activities, improving security over time.

The partnership between Vonage and AWS embarks on an innovative journey in the marketplace with Fraud Defender.

Transcripts

hi Chad hrin principal Solutions

architect with AWS I focus on

Communications as a service CX and

generative AI today I'm proud to

introduce our partner Vonage who's a

very Innovative Progressive partner

focusing on a solution called fraud

Defender Vonage offers communication

platform as a service solutions and

fraud Defender is actually taking

advantage of a lot of AWS AIML and

generative AI Solutions today you'll see

those moving parts and the architecture

how it comes together I'll turn it over

to the Vonage team hi my name is bort

from Vonage part of Ericson and we're

here today to walk through to the the

value that we're bringing as this

incredible Innovative and actually

industry first type of solution and so I

think what would be great is Mory if you

could help us with starting off with

what are the challenges that we have

within our

Marketplace yeah thank you Courtney my

name is morsey Chu I'm director of

global alliances at Vonage and the

Vonage fraud protection solution

actually addresses a large number of

challenges that organizations encounter

in terms of Fraud and fraud protection

as a first step CH organizations are

encountering challenges in terms of

identity theft and in terms of account

takeover this means that frosters today

are capable of combining separate

Technologies including

telecommunications capabilities in order

to hack and social engineer into

customer accounts when a type of attack

like account takeover happens this

causes usually a lot of financial losses

for the customer and it provokes also uh

uh brand damage to the

organization we also help organization

to protect themselves from

telecommunications fraud which is which

happens when a frauders try to abuse use

actually for example verification

processes to benefit uh their Financial

Health so this is the number of

challenges that we encounter

today but I think what's really

important that we want to highlight here

is as we take this approach with the

generative AI as well as the partnership

that we have with our Vonage team here

is it's really about how we're

protecting against the fraud and

identity fraud that's happening and how

that then connects with network apis our

Sim swap even our Cass Solutions and

then how it then progresses into account

takeover which is really very

Troublesome for not only the consumer

but also businesses fintech in

particular and then finally how this is

going to really help in curbing the

financial losses that are being seen

right now with fraudulent activities

that are happening every day constantly

so why don't we go into taking a look at

the architecture absolutely so the

architecture as you can see is a

completely Cloud native and serverless

architecture that is deployed on AWS

customers can purchase this through the

AWS Marketplace and they can benefit

from all the advantages of this solution

and it is easily customizable what we

can see in the architecture is a number

of components first of all there is the

application which is basically the

vehicle through which the uh user

interacts with the banking services or

with the financial institution there's

the application back end which is

located here and then we find the Vonage

fraud protection Suite of apis so these

apis are based on the Camara Network

apis and what they uh what they enable

to do is to get realtime insights about

what is happening on the network with

these insights we collect data that can

help us identify whether a transaction

or an authentication process is

fraudulent or not and so the information

that we collect we is also uh fed into

Amazon Bedrock Amazon Bedrock is a

generative Ai and llm model that allows

us basically to have uh a feedback on

those transactions and whether those

transactions were uh actually fraudulent

or not and then we use the feedback from

Amazon bedrock in order to feed our

fraud Defender and that allows to close

the loop between what is happening in

real time and how the system should

respond to fraudulent attacks or in the

other um uh better case where the user

is legitimate and therefore Grant access

to that user so what we can see here is

the full process that starts by the user

asking to register or to authenticate to

for example a banking application there

are multiple steps that happen all of

this happened in a number of mult 100

milliseconds but the process is actually

uh behind it is actually uh a

multi-layer approach of you know making

checks about the fraud about the

fraudulent capabilities and fraudulent

insights for example when a user tries

to connect and Supply their email

address password and uh their phone

number we start checking whether this

phone number is a fraudulent number

based on the fraud Defender uh knowledge

base um and then we can um move on with

checking other databases and other

Telecom Network as well so the csps also

have information about that we combine

all of this so that the bank has a

complete knowledge about that uh

fraudulent number we do also Sim swap

checks the Sim swap check enables us to

say if there is maybe a telecom fraud

happening in order to take over the uh

uh phone number of the user and that

basically can give us information

whether a transaction for example like a

password reset request is this a

legitimate access or is this or is this

a fraudster trying to take over an

account of a legitimate user we do also

silent authentication and Silent

authentication is a a very secure

process that can replace uh onetime

passwords for example it prevents social

engineering and at the same time it

provides a better guarantee about the

phone line that we are using in order to

authenticate the user the device

location can enable us to verify that

the user who is for example connecting

to a bank account from a certain country

or a certain region is actually also

located in that region and this is based

on the network information the real-time

Network information that the Telecom

operator can provide us through the

camar network API and finally we can

proceed with Amazon recognition Amazon

recognition does the user authentication

based on facial recognition and this is

a powerful process because it has a

liveliness check liveliness check is an

important piece of the process because

froster can now use social media for

example to recover photos and videos of

the user and try to use that for the

authentication process so Mory that

means that there's actually three

Network apis that we're talking about

which is number verification Sim Swap

and verify location absolutely so these

are network apis that are available

today in a certain number of countries

and uh they provide us with those

real-time insights that otherwise we

would not be able to get them uh through

the uh telecommunications

Network so really what this also does is

it takes away the need for the user to

have to sit and try to play with a

password anymore so what you're saying

is is that you don't have to remember

your passwords anymore cuz I'm terrible

at that and so now this is just

automatic yeah absolutely so what is

happening with this solution usually

when we increase the protection and the

the fraud protection in a certain uh

workload this usually adds friction to

the user and what is happening with this

process is exactly the opposite we're

increasing the capabilities of frog

protection and at the same time with

mechanisms like silent authentication

and uh facial recognition we actually

reduce the friction and we increase the

friction only when it is really needed

when we have insights that are telling

us that this could be a really a

fraudulent attempt to take over the

account of our legitimate user okay but

one final thing which makes I think it

really cool is is that this Amazon

recognition and the whole generate AI

all of a sudden starts making the

products even smarter more intelligent

more predictive which equals more secure

100% in order to do that we have fraud

Defender which is a vantage API and a

service that helps hundreds of thousands

of customers to protect their Network

and we use the insights that are coming

from fraud Defender to protect our

customers and what we do with the

generative AI we are capable of closing

the loop between the authentication

process and fraudulent activities and

suspicions and we feed that back to the

fraud to fraud Defender and that

increases the security over time and

also we fine-tune the parameters and we

make sure that the level of friction is

corresponds exactly to the risk that we

are perceiving on the

network so again thank you so much for

taking the time to be with us and learn

about Network apis the value that

Network apis and the generate Ai and our

partnership that we're we are embarking

upon with AWS it's incredibly exciting

we have so much that's coming and

there's so much more ahead of us but

this is just the first big step of

something that's incredibly Innovative

out there in the marketplace

[Music]